Abstract
Abstract. The automatic reconstruction of 3D building models with complex roof shapes is still an active area of research. In this paper we present a novel approach for local and global regularization rules that integrate building knowledge to improve both the shape of the reconstructed building models and their accuracy. These rules are defined for the planar half-space representation of our models and emphasize the presence of symmetries, co-planarity, parallelism, and orthogonality. By not adjusting building features separately (e.g. ridges, eaves, etc.) we are able to handle more than one feature at a time without considering dependencies between different features. Additionally, we present a new method for reconstructing buildings with concave outlines using half-spaces that avoids the need to partition the models into smaller convex parts. We present both extensions in the context of a fully automatic feature-driven 3D building reconstruction approach where the whole process is suited for processing large urban areas with complex building roofs.
Highlights
The automatic reconstruction of 3D building models with detailed roof structures in urban areas that are suitable for analysis tasks and are not restricted to visualization purposes is still an active research area
These rules take into account advanced building knowledge about local and global building regularities, symmetries, coplanarity, parallelisms and orthogonality
By defining them directly on half-spaces we are able to adjust more than one building feature at the same time without considering the interdependencies of building features
Summary
The automatic reconstruction of 3D building models with detailed roof structures in urban areas that are suitable for analysis tasks and are not restricted to visualization purposes is still an active research area. In model-driven approaches building templates are chosen from a predefined catalog and adapted by their parameters to best fit their roof shapes to the given data (Maas and Vosselman, 1999). To avoid extensive catalogs for the reconstruction of complex building models data-driven methods have been developed These bottom-up approaches start with a segmentation to obtain sets of points which usually determine planar regions representing roof faces. We extend the previously suggested feature recognition step by allowing certain parts of some low-level features to have in certain cases other parameter values than the rest of the feature The advantage of this extension is that we are able to reconstruct high-level features like gable or hip roofs with concave building outlines by sets of half-spaces. We can perform our regularization rules either on certain parts of a low-level feature or on all its parts at the same time
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
